CROSS CORRELATION FUNCTION FOR IDENTIFYING THE ORDO OF TIME-SERIES REGRESSION IN PONTIANAK^S COVID CASES Nur^ainul Miftahul Huda- Yundari
Universitas Tanjungpura
Abstract
In the time series model, identifying the order used to determine the time series model is based on the autocorrelation and partial autocorrelation functions, known as ACF and PACF. One of the functions that can also be used as identification to determine the order of the model is the cross-correlation function, known as CCF. In this study, CCF is used to identify the order of influence of an independent variable on the dependent variable in the time series regression model. Usually, in previous studies, the time series regression model used is the ARIMAX(p,d,q) model, where the influence of the independent variable on the dependent variable lies in the same time lag, namely the time lag t. In reality, cases of independent variables are often encountered, affecting the same time lag, but the previous time lag also affects the dependent variable. In this study, ACF was used to identify autoregressive order (influence of dependent variable with a different time lag), and CCF was used to identify regression order (influence of independent variable to the dependent variable with a different time lag). The data used in this study is data on positive confirmed cases of Covid, suspect, and Rapid Diagnostic Test (RDT) in Pontianak City from January 1, 2021 - May 23, 2021. There are two models, with each dependent variable being a positive confirmed case and a suspect. Positively confirmed cases are thought to be influenced by the number of RDTs carried out, while Suspect cases are thought to be influenced by the number of positive confirmed cases. In the literature, these variables affect. However, in this study, it will be seen how significant the influence is statistically, and the final output is to see the prediction results of the two models for the following three time periods
